A correlation of molecular flexibility with ring size is also possible in both types of ethers I11 and IV. Compounds V, VI, and VII, which differ only in number of ring difluoromethylene groups (from four- to six-membered rings), vary in order of 0°F. viscosity in ring-size order of 6 >> 4 > 5. The same relationship in the Type IV ethers establishes the better fluid properties of perfluorinated cyclopentane and cyclobutane systems over cyclohexane systems. The thermal stabilities of most of these ethers averaged about 500°F., although V and VI1 ethers had stabilities above 550" F., with the high fluorine content of both ether types resulting in good fire resistance properties. The perfluorocyclobutene trimer, XV, was an extremely mobile fluid with a viscosity of 117 cs. a t -65"F., compared to the acyclic n-CI2Fz6, which is a solid a t 167" F. The thermal stability of XV was above 900" F., as measured by a pressure isoteniscope. LITERATURE CITED
Blake, E.S., Edwards, J.W., Hammann, W.C., Reichard, T.E., Ort, M.O., J. CHEM.ENG.DATA6, 87 (1961). (2) Brice, T.J., Cohn, R.I., J . A m . Chem. SOC.75, 2921 (1953). (3) Clayton, A.B., Roylance, J., Sayers, D.R., Stephens, R., Tatlow, J.C., J . Chem. SOC.1965, p. 7358. (1)
(4) Faurote, P.D., Henderson, C.M., Murphy, C.M., O'Rear, J.G., Ravner, H., Id. Eng. Chem. 48,445 (1956). (5) Gash, V.W., Bauer, D.J., J . 0%.Chem. 31, 3602 (1966). (6) Henne, A.L., Latif, K.A., J . I d . Chem. SOC.30, 809 (1953). (7) McKee, E.T., Crain, D.L., Crain, R.D., Betohlav, L.R., Braendlin, H.P., J . Am. Chem. SOC.84, 3557 (1962). (8) Park, J.D., Dick, J.R., Lacher, J.R., J . 0%. Chem. 28, 1154 (1963). (9) Park, J.D., Lacher, J.R., Dick, J.R., Ibid., 31, 1116 (1966). (10) Park, J.D., Sharrah, M.L., Lacher, J.R., J . Am. Chem. SOC. 71, 2337 (1949). (11) Park, J.D., Snow, C.M., Lacher, J.R., Ibid., 73, 2342 (1951). (12) Park, J.D., Wilson, L.H., Lacher, J.R., J . Org. Chem. 28, 1008 (1963). (13) Pruett, R.L., Bahner, C.T., Smith, H.A., J . Am. Chem. SOC. 74, 1638 (1952). (14) Rapp, K.E., Pruett, R.L., Barr, J.T., Bahner, C.T., Gibson, J.T., Lafferty, R.H., Jr., Ibid., 72, 3642, 3646, 4480 (1950). (15) Shepard, R.A., Lessoff, H., Domijan, J.D., Hilton, D.B., Finnegan, T.F., J . 0%.Chem. 23, 2011 (1958). (16) Stockel, R.F., Beachem, M.T., Megson, F.H., Can. J . Chem. 42, 2880 (1964). (17) West, R., Nin, H.Y., Ito, M., J . Am. Chem. SOC.85, 2584 (1963). RECEIVED for review September 3, 1968. Accepted March 26, 1969.
Copper Chromite Reduction of Ethyl Pinonate J. B. LEWIS and G. W. HEDRICK Naval Stores Laboratory, Olustee, Fla.
R. L. SETTINE University of Mississippi, University, Miss. The effect of copper chromite catalyst on the hydrogenation of ethyl pinonate, ethyl 3-acetyl-2,2-dimethylcyclobutaneacetate was to reduce the carbonyl at 140' C., giving ethyl pinolate, some of which pyrolyzed and gave 3-vinyl- and J-ethylidienecyclobutane derivatives before the carboxyl function was reduced to alcohol. The result was a mixture of 2,2-dimethyl-3-ethylcyclobutaneethanol and 2,2-dirnethyl-3-( 2-hydroxyethy1)-cyclobutaneethanol. Because the yield of mono alcohol was about 38% and of glycol only 34%, copper chromite reduction of ethyl pinonate is considered impractical.
A NUMBER of papers have emanated from this labora-
1.4526 ( 4 ) ] ,prepared by the esterification of cis-dl-pinonic tory on the chemistry of dl-pinonic acid (I,) and dl-pinolic acid with ethanol in toluene using p-toluenesulfonic acid acid (II.A). This report is closely related to some of the as a catalyst, was reduced with lithium aluminum hydride earlier work and was prompted by a request for a large CHz CH? sample of the glycol, 2,2-dimethyl-3-(2-hydroxyethyl)-cyclobutaneethanol (111) for evaluation as an intermediate for polyurethanes. Accordingly, a study of the copper chromite reduction of ethyl pinonate (I,) was made as a preparative method for this glycol. When I, was reduced, the reduction product was a mixture of 111, and 2,2-dimethyl-3-ethylcyclobutaneethanol(IV) . The composition of each was obtained by comparing 111, and IV, with materials made by alternate syntheses, III,, 111,, and IV,. EXPERIMENTAL
Boiling points, "C., are uncorrected. The carbon and hydrogen analyses were made by Galbraith Laboratories, Inc., Knoxville, Tenn. The GLPC data were obtained from (A) a column Yq inch by 15 feet, packed with 30% Carbowax 20M on Chromosorb W , or (B) a column 3/16 inch by 12 feet, packed with 2OCC Carbowax 20M on Gas Chrom
Q.
IIIA FROM COPPER CHROMITE REDUCTION OF Ig IIIB FROM THE LAH REDUCTION OF Ig
IIIc
IVA
COPPER CHROMITE REDUCTION OF
IVB
IB
HYDROGENATION OF V
FROM THE L A H REDUCTION OF 1 1 4
CH2'CH
2,2-Dimethyl-3-( 2-hydroxyethy1)cyclobutaneethanol (111). A.
Ethyl pinonate, I,, 360 grams (1.65 mole) VOL. 14, No. 3, JULY 1969
7;
1.4526 [lit.
V
V I
40 1
(LAH) in accordance with Park et al. ( 2 ) , except that decomposition of the LAH complex was accomplished by dropwise addition of water so that the solids agglomerated and settled quickly when stirring was interrupted. Treatment with mineral acid was unnecessary. The product was isolated by filtering to remove the sludge, washing the sludge with ether, evaporating ether, and distilling, 210 grams of I,, 73.9%. 7; 1.4754. Anal. Calcd., CloHraOz: C, 69.72; H , 11.70. Found: C, 69.93, H , 11.90. The chromatogram from the GLPC analyses, column A or B, had four peaks with relative peak areas of 34.4, 7.6, 2.5 and 1.0. The GLPC chromatogram of I,, column A, had two peaks with relative areas of 4 to 1. B. Solid pinolic acid (II,), m.p. 104°C. from water, 2.0 grams (0.011 mole) was reduced similarly with LAH and gave 1.65 gram of product, 100%. The GLPC chromatogram, column A, had two peaks with relative areas of 6.8 to 1. These had the same retention times as peaks 1 and 2 above. C. Copper Chromite Reduction of II. Ethyl pinonate, I,, 586 grams (2.8 moles), was reduced neat in a rockingtype pressure reactor a t 140°C. with 4 to 5000 p.s.i. of hydrogen and 58 grams of barium promoted copper chromite powdered catalyst (Girdler G-22). A sample was taken, dissolved in benzene, and filtered to remove the catalyst. After removal of solvent, analyses of the residue by GLPC and by infrared spectroscopy indicated the presence of a substantial amount of 11,. The entire batch was worked up like the sample and recharged in the pressure reactor with 50 grams of fresh catalyst, hydrogen added a t 3,000 p.s.i., and heated to 275°C. After about five hours, the reactor was allowed to cool and the product dissolved in benzene, filtered, and distilled. Two batches were combined, giving 372 grams of slightly impure IV.&, 38.2%, b.p. 92-96"C., 3.0 mm., 7; 1.4497, and 331 grams of 111,, 34.470, b.p. 136-138" C., 2.0 mm. Hydroxyl equivalent calcd. for C10H2002: 86.13. Found: 88.1, 7: 1.4699. The composition of the lower-boiling fraction was confirmed by comparing the GLPC chromatogram, column B, and infrared spectrum with that of IV, obtained from hydrogenation of the vinyl derivative, V (below). The GLPC chromatogram, column A or B, for the higherboiling fraction had the same four peaks as the glycol obtained by the LAH reduction of ethyl pinonate above. Since the relative peak areas (1.67, 1, 8, and 1.67) were different, it is assumed the same four isomers are present, but in different amounts. The glycols 111,, 111,, and 111, had identical infrared spectra. 2,2-Dimethyl-3-vinylcyclobutaneethanol ( V ) . Ethyl 2,2-dimethyl-3-vinylcyclobutaneacetate (VI)4, 19.4 grams (0.1 mole), was reduced with 6 grams LAH, as above. A colorless slightly viscous liquid was obtained, 14 grams, 91%, b.p. 78-79°C. a t 2 mm., 7; 1.4646. There were strong vinyl bands in the infrared a t 908, 1649, and 3010 cm.-'. Anal. Calcd. for Cl0HI80:C, 77.68; H , 11.77. Found: C, 78.01; H , 11.47. 2,2-Dimethyl-3-ethylcyclobutaneethanol (IV,). The vinyl compound V, 2.0 grams, in ethanol was reduced with hydrogen, 30 p s i . , using platinum oxide catalyst. Isolating the product gave 1.93 grams, 7; 1.4491. DISCUSSION
Ethyl dl-pinonate made by direct esterification of cis I, is a mixture of cis and trans isomers in a ratio of 4 to 1. Similarly, dl-pinolic acid used in this work was a mixture of cis and trans isomers. Thoi ( 5 ) , Trave ( 6 ) , and other workers have assigned the name of cis-dl-pinonic and cis-dl-pinolic acids to the high-melting isomers (105" C. for both pure acids). The lower-melting or liquid materials were assigned t o the trans configuration. The relative amounts of the isomers of esters of either of the acids 402
used in this work were readily discernible by GLPC. Nomenclature of these authors is used in this report. Both I, and 11, were reduced with LAH t o obtain 111, and 111, for comparison with 111,. The product 111, had two peaks in its GLPC chromatogram, and must have been a mixture of cis and trans isomers. The chromatogram from 111, had four peaks, two of which had the same retention times as the isomers of 111,. Accordingly, the glycol 111, must have contained the cis and trans isomers of 111,, and it is reasonable to expect the presence of two diastereomers isomeric with the above cis and trans isomers. No subsequent work, however, has been done to establish this. The chromatogram of 111, had four peaks with retention times identical to those of 111,, but the relative composition as represented by the peak areas of the chromatogram was somewhat different. The elemental analyses, hydroxyl equivalent, and infrared spectral analyses are in support of the structure 111, as assigned. Park et al. (2) used LAH reduction of a number of similar materials giving a number of hydroxyalkyl and olefinic substituted gem-dimethylcyclobutanes without complications from isomerization or rearrangement. Park and coworkers also studied the pyrolysis and rearrangement of 11, ( 3 ) . Data in this report and accumulated data in this laboratory show that products of rearrangement would be discernible by GLPC. From the evidence presented, there seems to be little doubt that 111, and 111, are mixtures of glycols as described. The mono-alcohol IV, was prepared by hydrogenation of the vinylcyclobutane derivative, V, which in turn was obtained from LAH reduction of VI prepared and described by Lewis and his collaborators ( I ) . Since the ester, VI, was a cis-trans mixture, IV, and V were also mixtures. Comparison of IV, from the copper-chromite reduction with IV, by GLPC and infrared spectroscopy leaves little doubt that IV, is a cis-trans mixture like IV,. The formation of IV, is believed to be simple and direct. During copper-chromite reduction of I,, adsorption of hydrogen was quite rapid a t 140°C. Letting the reaction cool to stop the reaction and examining a sample by GLPC and infrared spectroscopy showed the presence of an appreciable amount of 11,. Hydrogenation was continued. No appreciable amount of hydrogen up-take was observed until the temperature reached 250°C. The ketone was reduced a t about 140°C., giving ethyl pinolate, some of which pyrolyzed a t the higher reaction temperature. Hydrogenolysis rather than pyrolysis is not precluded; however, the conditions were practically the same as those described by Park et al. for the pyrolysis of pinolic acid (3). If pyrolysis occurred, the resulting 3-vinyl- and 3-ethylidenecyclobutane derivatives would hydrogenate, resulting in IV,. In view of the results obtained, use of a copper-chromite reduction for the preparation of 111, on a large scale seems impractical because of the susceptibility of pinolic acid to pyrolysis. LITERATURE CITED (1) Lewis, J.B., Settine, R.L., Mayne, R.Y., Hedrick, G.W., J . Med. Chem. 11, 955-7 (1968). Hedrick, (2) Park, J.D., Allphin, N.L., Jr., Choi, S., Settine, R.L.,
Eng. Chem. Prod. Res. Deuelop. 4, 149 (1965). G.W., Id. (3) Park, J.D., Settine, R.L., Hedrick, G.W., J . 0%. Chem. 27, 902 (1962). (4) Parkin, B.A., Jr., Hedrick, G.W., J . Org. Chem. 25, 1417 (1960). (5) Thoi, Le-Van, Ann. Chim. (Rome) 10, 35 (1931). (6) Trave, R., Chim. Ital. 85, 908 (1958). for review September 9, 1968. Accepted March 21, 1969. RECEIVED Mention of commercial products is for identification only and does not constitute endorsement by the U. S. Department of Agriculture over those of other manufacturers. JOURNAL OF CHEMICAL AND ENGINEERING DATA